JP7050334B2 - Optical RF spectrum analyzer - Google Patents

Description

Cross-reference to related applications This application claims priority from Australian Patent Provisional Application No. 2016903330 filed on August 22, 2016, the contents of which are incorporated herein by reference.

The present disclosure relates to an optical RF spectrum analyzer and a method for analyzing an input RF signal.

The increasing demand for the expansion of the functionality of electronic RF systems presents an unprecedented challenge of providing a system capable of recognizing microwave signals with high resolution and wide operating bandwidth. There are various photonic-assisted channelization approaches to report the presence of detected microwave signals and perform instantaneous spectral activity monitoring including parameter mixing, coherent optical frequency combs, and induced Brulian scattering. These schemes are based on the use of multiple light sources, which increases the complexity and cost of the system. Also, existing photonic-assisted channelized receivers are limited by the one-to-one relationship between optical filtering bandwidth and radio frequency (RF) measurement resolution, i.e., RF measurement resolution depends only on optical filtering bandwidth. do. For example, to have an RF measurement resolution of 20 MHz requires a highly selective optical filter with a 3 dB bandwidth of 20 MHz, which is very complex or even impossible to achieve in stereolithography.

This essential drawback is brought about by the spectral widening effect between adjacent frequency components caused by the limited bandwidth and the selectivity of the optical filter. This significantly reduces the system performance for accurately retaining the frequency information of the microwave signal. Also, signal processing using a silicon ring resonator is very interesting due to its small size and compatibility with CMOS manufacturing technology. To produce a highly selective optical filter with band response and narrow passband (less than 100 MHz) for photonic-assisted channelized receivers, the approach is for each gap and ring in the critical coupling state and coupling. Use a large number of rings with precise control of dimensions. Such complex technical processes severely limit the possibility of integrating channelized receivers into a single chip.

Optical RF spectrum analyzer
An optical modulator for modulating the input RF signal to the carrier frequency,
An optical spectrum weight that has a spectral weighting function for modifying a modulated optical signal and determines the frequency relationship between the spectral weighting function and the carrier frequency,
A frequency control module for correcting the frequency relationship between the spectral weighting function and the carrier frequency over time,
An optical sensor for sensing the modified optical signal over time and generating an RF signal over time,
It includes a signal restoration module for calculating the RF spectrum over time based on the RF signal.

By modifying the relationship between the spectral weighting function and the carrier frequency over time and then calculating the RF spectrum over time from the filtered RF signal, high spectral resolution is achieved even when the spectral weights are relatively wideband. It is an advantage to occur. This is an advantage over other methods that rely on very narrowband filters, which are difficult and / or expensive to manufacture and / or operate. The proposed method results in higher spectral resolution with reduced price / complexity and increased robustness.

The signal restoration module may be configured to perform deconvolution of the RF signal over time based on the spectral weighting function.

Deconvolution may be based on an analytic approximation of the spectral weighting function.

The signal restoration module may be configured to perform deconvolution of the RF signal over time by using the waveform of the RF signal over time as a frequency domain signal.

Using the waveform of an RF signal over time as a frequency domain signal creates a frequency axis associated with the RF signal over time based on the rate at which the frequency relationship between the spectral weighting function and the carrier frequency is modified over time. You may be prepared for that.

The spectral weights may be resonant. The resonance may be a ring oscillator.

The light modulator may include a laser light source for producing a laser at carrier frequency. The frequency control module may be for modifying the relationship between the spectral weighting function and the carrier frequency by changing the carrier frequency. Changing the carrier frequency may comprise performing a sweep over a frequency range.

Modifying the relationship between the spectral weighting function and the carrier frequency over time may be based on the rate of change per unit time, and the signal recovery module is for calculating the RF spectrum based on the rate of change. good.

The method for analyzing the input RF signal is
Modulating the input RF signal to the optical carrier frequency to generate a modulated optical signal,
Modifying a modulated optical signal by having a spectral weighting function and assigning a spectral weighting that determines the frequency relationship between the spectral weighting function and the carrier frequency.
Correcting the relationship between the spectral weighting function and the carrier frequency over time to generate the corrected optical signal over time,
Sensing the modified optical signal over time and generating an RF signal over time,
It comprises calculating the spectrum of the input RF signal over time based on the RF signal.

Computing the spectrum of the input RF signal may comprise performing deconvolution of the RF signal over time based on a spectral weighting function.

Deconvolution may be based on an analytic approximation of the spectral weighting function.

Performing deconvolution of the RF signal over time may comprise using the waveform of the RF signal as a frequency domain signal.

Using the waveform of an RF signal as a frequency domain signal comprises generating a frequency axis associated with the RF signal over time based on the rate at which the frequency relationship between the spectral weighting function and the carrier frequency is modified over time. It's okay.

Modifying the relationship between the spectral weighting function and the carrier frequency may comprise changing the carrier frequency of the laser source.

Changing the carrier frequency may comprise performing a sweep over a frequency range.

Modifying the relationship between the spectral weighting function and the carrier frequency over time may be based on the rate of change per unit time, and calculating the RF spectrum may be based on the rate of change.

A method for controlling RF signal analyzers,
The relationship between the carrier frequency and the spectral weighting function that generates the modulator control signal to control the modulation of the input RF signal to the carrier frequency by the optical modulator and characterizes the spectral weights for modifying the modulated optical signal. To correct over time and
Receiving a sensed RF signal indicating the modified signal generated by the optical spectrum weights over time,
It comprises calculating the spectrum of the input RF signal over time based on the RF signal.

An example is illustrated with reference to the following figure.

The RF spectrum analyzer is shown. The method for RF spectrum analysis is shown. An example of a signal in the RF spectrum analyzer of FIG. 1 is schematically shown. An experimental device for photonic-assisted RF measurement is shown. The measured spectral response (solid line) and simulated spectral response (dotted line) of the optical filter are shown. Top microscopic image of a prefabricated silicon photonic single ring ad drop filter. Shows the measured and modeled output light intensity. The estimated RF signal frequency at 20 GHz is shown. Shows the measured and modeled output light intensity. The estimated RF signal when the input RF frequency is 0.5 GHz is shown. Shows the measured output light intensity. The estimated RF when the input RF frequency is 20 GHz and its power changes is shown. The inset is an enlarged estimation result. The upper solid line is 0 dBm RF power, the middle dotted line is -10 dBm RF power, and the lower solid line is -20 dBm RF power. The schematic diagram of the RF spectrum analyzer based on the filter bank is shown. The spectrum of the RF spectrum analyzer of FIG. 8 is shown. The schematic diagram of the RF spectrum analyzer based on the laser filter array is shown. The spectrum of the RF spectrum analyzer of FIG. 10 is shown. A schematic diagram of an RF spectrum analyzer based on an adjustable filter is shown. The spectrum of the RF spectrum analyzer of FIG. 12 is shown. The schematic diagram of the RF spectrum analyzer based on a dispersion element is shown. The spectrum of the RF spectrum analyzer of FIG. 14 is shown. The schematic diagram of the RF spectrum analyzer based on the RF multiplier is shown. The spectrum of the RF spectrum analyzer of FIG. 16 is shown.

The present disclosure breaks the one-to-one relationship between RF measurement resolution and optical bandwidth, provides high RF measurement resolution without increasing design and manufacturing complexity, and provides both amplitude and frequency information for microwave signals. It provides the technology to restore and enable on-chip RF frequency measurement systems.

FIG. 1 shows an RF spectrum analyzer 100 with a signal input 101, a modulator 102, a frequency control 103, such as a spectral weight 104 such as a filter, an optical field integrator 105, and a signal restoration module 106.

FIG. 2 shows a method 200 for RF spectrum analysis and FIG. 3 shows a signal flow 300 for RF spectrum analysis. The reference numbers in FIGS. 1, 2, and 3 correspond to each other in the sense that the reference numbers 101, 201, and 301 correspond to each other.

At the time of use, the signal input 101 receives the input signal 301, and the modulator 102 generates a modulated signal 302 having an optical sideband 202. The frequency control 103 maps the frequency signal 302 to the time domain 303, and the spectrum weight 104 assigns the spectrum weight 304 to the time domain signal 204. Finally, the frequency integrator 105 integrates the signal 205 to generate the integration time signal 305, and the signal restore module 106 restores the signal 206 to calculate the RF spectrum of the input signal 301.

The input 101 may be an antenna that receives a spectrally limited (narrowband) input signal. In the frequency spectrum of the input signal 301, the narrowband signal is shown at 310. For the sake of brevity, the narrowband signal 310 is assumed to be symmetric with respect to the center frequency 311. In other words, the center of the narrowband signal 310 is separated from the starting point of the frequency axis by the center frequency 311. However, keep in mind that the input signal may have any spectrum and in many cases the input spectrum is unknown.

Modulation by the modulator 102 in step 202 produces an upper sideband 320 and a lower sideband 321. As a result of the modulation, the centers of the upper sideband 320 and the lower sideband 321 are separated from the modulator frequency 322 by the center frequency 311. The center frequency 311 may be the optical frequency of the laser. Since the optical frequency is orders of magnitude higher than the radio frequency, the frequency axis is divided in FIG. An example of the RF center frequency 311 may be 20 GHz and an example of the carrier optical frequency 322 may be 195 THz.

The frequency control 103 maps the modulated signal 302 to the time domain by sweeping the carrier frequency 322 from low light wavelengths to high light wavelengths, for example from 1546.45 nm to 1547.25 nm. In short, this shifts the sidebands 320, 321 and the carrier 322 moves to the right over time as indicated by the time arrow 303. In one example, the sweep rate is 10 MHz per ms.

Most integrators are broadband and effectively integrate the entire spectrum of the modulated signal 302, so by applying an integrator 105, such as a photodiode, directly to the modulated and time-mapped signal 302, only a small amount of time has passed. Keep in mind that only target changes occur. Therefore, there is little or no beneficial change over time due to sidebands 320, 321 and carrier shifts. However, the spectral weight 104 is added to the signal before the integrator 105 is applied. The spectral weight may be optical resonance, and an example of frequency response is depicted in 304 of FIG. Adding a spectral weight basically means multiplying the frequency spectrum 302 at each frequency by the spectral weight 304 at that frequency and then integrating over the entire frequency range. This is also called convolution in the time domain and is multiplication in the frequency domain.

It can be seen that there is no overlap between the two signals at the relative positions of the modulated signal 302 and the spectral weight 304 shown in FIG. In other words, at least one of the signals is zero at all points on the frequency axis. As a result, the product of the two signals is also zero across the frequency axis. By integrating the zero signals, the zero output shown at time t ₁ 350 in the integrated time signal 305 is produced. Since the modulated signal 302 is shifted to the right 303 over time, the upper sideband 320 begins to move so as to overlap the spectral weight 304, and the product of the upper sideband and the spectral weight becomes non-zero. Overlapping increases over time, which is indicated by the first peak 351 of the integration time signal 305. When the upper sideband 320 deviates from the spectral weight 304, the integrated time signal 305 recedes to zero until the carrier 322 moves over the spectral weight 304 and results in a second peak 352. Similarly, the lower sideband moving over the spectral weight 304 results in a third peak 353.

In other words, the spectral overlap can be expressed as the frequency relationship between the spectral weighting function 304 and the carrier frequency 322. The frequency relationship is defined by the spectral weights in the sense that the spectral weights define the relative arrangement or alignment of the carrier frequency 322 and the spectral weighting function 304. It may include a fixed spectral weighting function, for example implemented by a fixed optical ring oscillator. In another example, the spectral weights may be adjustable, for example by adjusting an optical ring oscillator or using an optical processor. Regardless of the adjustment, the spectral weights determine the frequency relationship between the carrier frequency 322 and the spectral weighting function 304. The particular spectral relationship may be that the spectral weighting function is at a frequency significantly higher than the carrier, 10 GHz above the carrier, or above the frequency of the upper sideband 320 and above the carrier. Modifying the frequency relationship over time may include the sweeps described above and may result in the overlaps described above.

In one example, the three peaks 351, 352, and 353 are clearly separated as shown in the integrated time signal 305 of FIG. In this case, the restore module 106 can detect the three peaks 351, 352, and 353 as maximum values and measure the time t ₀ 354 between the upper sideband peak 351 and the carrier peak 352. This time t ₀ 354 indicates the time between the overlap of the upper sideband 320 and the spectral weight 304 and the overlap of the carrier wave 322 and the spectral weight. That is, the time t ₀ indicates the period during which the modulated signal spectrum is shifted from f ₀ . Since the shift speed, which is the sweep speed in Hz / s, is preset or known as an input to the frequency control 103, the frequency f ₀ can be calculated as f ₀ = speed × t ₀ . This indicates that the spectral weights 304 are in a relatively wide band and do not significantly reduce the spectral resolution of the output as long as the peaks 351, 352, and 353 are distinguishable.

However, in another example, the three peaks 351, 352, and 353 are difficult to distinguish by strongly overlapping peaks. In this case, the restore module 106 may apply a deconvolution algorithm to the time signal. However, it should be noted that many processes in signal processing are represented by the convolution of two time signals, such as an input signal and a filter response, which can be converted into multiplications in frequency space. In contrast, here the convolution occurs in the frequency space between the modulated signal 302 and the spectral weighted spectrum 304, but the output is the time domain integrated time signal 305. However, it is possible to replace the time axis of the integrated time signal 305 with a frequency axis that follows the sweep speed. For example, if the sweep rate is 10 MHz per second, the restore module multiplies the time value on the time axis of the integrated time signal 305 by 10 MHz / s, converts them to frequencies, and applies an offset if applicable. It's okay. The waveform of the integrated time signal 305 can then be passed to the deconvolution algorithm as a frequency domain signal using the determined frequency axis to calculate the estimated value of the original signal 302. Deconvolution is, for example, scipy. signal. It can be executed by the processor of the computing system by executing the Python code in the deconvolve package. In another example, the FPGA performs deconvolution.

The accuracy of the deconvolution result can be enhanced by using the actual spectrum of spectrum weight 204 or an approximation thereof. One approximation may be a Gaussian distribution. A more accurate approximation may be Dirac impulse.

In short, a spectral analysis system can be implemented by five signal processing steps in both the optical and electrical regions, including optical sideband (OSB) generation, frequency time mapping, spectral weighting, optical field integration, and signal restoration. First, the input RF signal is applied to the optical carrier wave generated by the light source via electro-optical modulation, thereby realizing optical sideband generation. Second, the optical sideband information in the optical spectrum is transformed into a waveform in the time domain via frequency time domain mapping, where the intensity profile becomes a scaled replica of the optical spectrum. Third, the time-varying waveform is transmitted through a spectral weighting module, such as an optical filter, where its intensity at the output at each time instance is weighted according to the contributing frequency component. Fourth, optical field integrators, such as photodetectors or wattmeters, are utilized to measure accurate light intensity at each time instance. Fifth, a signal restoration module based on the deconvolution theorem is used to reconstruct the input RF signal spectrum from the light intensity measured by the optical field integrator.

Applications for this technology include:
・ Scanning receiver ・ RF spectrum analysis ・ Optical spectrum analysis ・ Photonic signal processing ・ Integrated photonic chip ・ Radar

To test the concept of deconvolution, it is based on the apparatus 400 shown in FIG. 4, which comprises an input 401 for providing an unknown RF input signal modulated by an adjustable laser light source 403 to the optical sideband generator 402. The experiment was conducted. It should be noted that the possible frequency measurement range is largely limited by the optical sideband generator used in the system. Current developments in light light modulators provide modulation bandwidths in excess of 100 GHz, which can facilitate systems using the proposed ultra-wideband RF spectral analysis. The modulated signal is fed to a single ring resonator 404 that acts as a spectral weight and is connected to an optical field integrator 405 that provides the signal to the signal restoration module 405. In the experiment, an adjustable laser (Keysight) 403 with a constant sweep rate of 80 nm / s was swept from 1546.45 nm to 1547.25 nm. This results in a total sweep period of 4000 points and about 10 ms, where the time interval of 1 microsecond in the time domain corresponds to a frequency change of 10 MHz. For other applications, such as optical spectrum analysis requiring wideband measurements, the sweep range may be further expanded, eg, the C + L band. The optical filter 404 for providing spectral weighting in the system 400 is based on a single ring resonator centered at 1546.85 nm with a 3 dB bandwidth of approximately 1.4 GHz. The measured light spectrum of the filter is shown in FIG. 5a.

First, an input microwave signal 401 with an RF frequency of 20 GHz was used in the test. FIG. 5c shows the light intensity waveform measured at the output of the optical field integrator 405, which was an optical wattmeter in the experiment. The results show the mapping of light power intensity to the time domain, the effect of OSB generation 402, and the weighting of light intensity. The frequency of the RF signal estimated at the output of the signal restoration module 406 is shown in FIG. 5d. The restoration of the input signal is performed by the inverse transformation by the existing signal deconvolution algorithm. For example, blind deconvolution methods for image and audio processing applications may be adapted and used in the restore module.

The measurement error is less than 25 MHz, which is significantly smaller than the filter bandwidth of 1.4 GHz.

6a and 6b show measurements when the input microwave frequency is reduced to 0.5 GHz, which is significantly smaller than the 3 dB bandwidth of the optical filter. As can be seen from FIG. 6a, the output light intensity of the photopower meter spreads towards a single peak, but the frequency of the RF signal can be accurately identified by the new technique shown in FIG. 6b.

7a and 7b show the measured light intensity wavelength and the estimated modulation signal, respectively, when the input RF frequency is fixed at 20 GHz and the power varies. As illustrated, this technique fully restores both the amplitude and frequency of the signal. The estimated amplitude errors are 0.72, 0.126, and 0.06 dB at input RF powers of 0 dBm, -10 dBm, and -20 dBm, respectively.

The combination of the frequency-time mapping module and spectrum weighting module with the optical field integrator and signal restoration module allows for high resolution input RF signal spectrum reconstruction and significantly reduces design and manufacturing complexity. Will be done. This technique improves resolution and operating bandwidth measurements, breaks the one-to-one relationship between RF measurement resolution and optical bandwidth, restores both amplitude and frequency information of microwave signals, and on-chip RF. It also enables a frequency measurement system. The proposed solution for microwave frequency measurement is less susceptible to wavelength drift of light sources and devices because the optical carrier and sideband information provide a self-reference function.

As mentioned above, deconvolution can be performed by the processor of the computer system, for example the restore module 106. The processor may also perform the control task of the spectrum analyzer 100, for example, under the instruction of software code stored in the program memory connected to the processor. These control tasks control the modulator to control the modulation of the input RF signal 101 to the carrier frequency 322 by the optical modulator 102 in order to modify the relationship between the spectral weighting function 104 and the carrier frequency 322 over time. It may include generating a signal. This basically means that there is another connection in FIG. 1 between the restore module 106 and the frequency control 103. The processor also receives a sensed RF signal 305 over time, indicating the correction signal generated by the optical spectrum weight 104. The processor then calculates the spectrum of the input RF signal over time based on the RF signal, for example by applying a deconvolution algorithm to the RF signal over time, as described above.

Further Embodiment

The following description provides a further embodiment. In general, the last two digits of a reference number indicate the corresponding feature, meaning that, for example, the RF signal 801 in FIG. 8 corresponds to the RF signal 101 in FIG.

FIG. 8 shows a schematic diagram of another method for optical RF spectral analysis using the filter bank 804 to modify the relationship between the spectral weighting function 804 and the carrier frequency 803 without changing the carrier frequency 803. The modulated optical signal is generated by modulating the target RF signal 801 from the laser 803 into a fixed wavelength carrier signal. This is then evenly divided and transmitted to the bank of the narrowband optical filter 804 having the same line shape. Each individual filter is centered on an individual frequency (f ₁ , f ₂ , ..., F _N ) with a fixed interval. The relationship between the carrier and the individual filters changes over time, as described above, for example by sweeping the carrier frequency. The use of multiple filters as shown in FIG. 8 reduces the time required for analysis. For example, when using four filters, it takes only one-fourth of the analysis time of one filter.

FIG. 9 shows how each split portion of the modulated signal is multiplied by the shifted spectral weights. The weighted spectrum of the modulated signal is then transmitted to the photodiode 805, where integration is performed over the entire spectrum of the modulated signal. In the signal recovery element 806, the power output of the photodiode is sewn together and deconvolved by the filter line shape, resulting in a real-time spectrum of the input RF signal.

FIG. 10 shows another example 1000 using a filter bank 1004 to implement an RF spectrum analyzer. The laser array 1003 is used to generate multiple carrier signals at various frequencies (fc ₁ , fc ₂ , ..., Fc _2N ). The carrier signal is then injected into the modulator 1002. The RF information is mapped to all carrier signals, resulting in 2N copies of the modulated signal centered on fc ₁ , fc ₂ , ..., fc _2N , respectively. This is then evenly divided and transmitted to the bank of the narrowband optical filter 1004 having the same linear shape. By adjusting the center frequency of the filter in the filter bank 1004, the relative distance between the nth (1 ≦ n ≦ N) copy of the modulated signal and the filter position is nΔf, and each filter output is the shifted spectrum. It is the product of the weighting function and the modulated signal.

FIG. 11 shows the spectrum of an RF spectrum analyzer based on a laser filter array.

Relationships between spectral weighting functions can also be modified over time by using adjustable optical filters. FIG. 12 shows a schematic diagram of an optical RF spectral analysis system 1200 based on a laser light source 1203 that produces a fixed wavelength carrier, modulator 1202, adjustable optical filter 1204, photodiode 1205, and signal recovery element 1206. By varying the center frequency of the adjustable filter 1204 with a fixed velocity factor, the optical filters are placed in different positions at each time instance (t ₁ , t ₂ , ..., T _N ) as shown in FIG. Yes, sweeping over the frequency range can be achieved.

FIG. 14 shows another example. As shown in FIG. 14, the laser array 1403 produces a plurality of carrier signals with fixed frequency intervals (fc ₁ , fc ₂ , ..., Fc _N ). The optical carrier signal is injected into the modulator 1402 to obtain N copies of the modulated optical signal 1401 centered on fc ₁ , fc ₂ , ..., Fc _N , after which the output of the modulator is delayed. It is transmitted to an element (for example, a distributed delay line). The time delay provided by the delay element is optical frequency dependent, indicating that each copy of the modulated signal arrives at the filter 1404 at various times. When a distributed device with a linear delay slope is used as the optical delay line 1410 in the system, there is a constant time delay between n copies of the modulated signal and n + 1 (1 ≤ n ≤ N) copies. Is done. Therefore, the relationship between the spectral weighting function over time and the RF spectrum is controlled by the dispersion characteristics of the dispersion element. FIG. 15 shows the corresponding spectra.

FIG. 16 shows a method in which the RF signal 1601 can be first transmitted to the frequency multiplier 1610 and then to the modulator 1602, rather than directly applying the input RF signal 1601 to the carrier signal generated by the laser light source 1603. Is shown. For example, a frequency doubler may double the input RF frequency as shown in FIG. 17, so that the relative spacing between the carrier signal and the sideband is also optical before being injected into the spectral weighting module. It doubles in the frequency domain. By varying the ratio of the multiplier 1610, the sidebands of the modulated signal are assigned to different positions in each time instance. Therefore, sweeping over the frequency range can be achieved.

As will be appreciated by those skilled in the art, a number of modifications and / or modifications can be made to the embodiments described above without departing from the broad general scope of the present disclosure. Therefore, this embodiment should be regarded as exemplary rather than limiting in all respects.

Claims (22)

An optical modulator for modulating the input RF signal to the carrier frequency to generate a modulated light signal,
A spectrum weighting module having a spectrum weighting function for modifying the modulated optical signal and determining a frequency relationship between the spectrum weighting function and the carrier frequency, and a spectrum weighting module.
The frequency relationship between the spectral weighting function and the carrier frequency is modified over time by performing a sweep over the frequency range to modify the carrier frequency and the modulated optical signal, and the modified optical signal is modified over time. With a frequency control module to generate,
An optical sensor for sensing the modified optical signal over time and generating an RF signal based on the modified optical signal over time.
An optical RF spectrum analyzer including a signal restoration module for calculating an RF spectrum over time based on the RF signal. The spectrum analyzer according to claim 1, wherein the signal restoration module is configured to perform deconvolution of the RF signal over time based on the spectral weighting function. The spectrum analyzer according to claim 2, wherein the deconvolution is based on an analytical approximation of the spectral weighting function. The signal restoration module is configured to execute the deconvolution of the RF signal over time by using the waveform of the RF signal as a frequency domain signal over time, according to claim 2 or 3. Described spectrum analyzer. Using the waveform of the RF signal over time as a frequency domain signal creates a frequency axis associated with the RF signal based on the rate at which the frequency relationship between the spectral weighting function and the carrier frequency is corrected over time. The spectral analyzer according to claim 4, further comprising producing over time. The spectrum analyzer according to any one of claims 1 to 5, wherein the spectrum weight module is a resonator or a ring oscillator. The spectrum analyzer according to any one of claims 1 to 6, wherein the light modulator includes a laser light source for generating a laser at the carrier frequency. Correcting the frequency relationship between the spectrum weighting function and the carrier frequency over time is based on the rate of change per unit time, and the signal restoration module calculates the RF spectrum based on the rate of change. The spectrum analyzer according to any one of claims 1 to 7. A bank of spectral weighting modules, each having a spectral weighting function for modifying the modulated optical signal, and each defining the frequency relationship between the spectral weighting function and the carrier frequency.
Further equipped with an array of optical sensors for sensing the modified optical signal over time and generating multiple RF signals over time.
The spectrum analyzer according to any one of claims 1 to 8, wherein the signal restoration module calculates the RF spectrum over time based on the RF signal. A method for analyzing input RF signals,
Modulating the input RF signal to the optical carrier frequency to generate a modulated optical signal,
Modifying the modulated optical signal by using a spectral weighting module that applies a spectral weighting function, the spectral weighting module defining a frequency relationship between the spectral weighting function and the optical carrier frequency.
The frequency relationship between the spectral weighting function and the optical carrier frequency is modified over time by performing a sweep over the frequency range to modify the optical carrier frequency and the modulated optical signal, and the modified optical signal is modified over time. To generate
To detect the modified optical signal over time and generate an RF signal over time based on the modified optical signal.
A method comprising calculating the RF spectrum of the input RF signal over time based on the RF signal. 10. The method of claim 10, wherein calculating the RF spectrum of the input RF signal comprises performing deconvolution of the RF signal over time based on the spectrum weighting function. The method of claim 11, wherein the deconvolution is based on an analytical approximation of the spectral weighting function. The method of claim 11 or 12, wherein performing the deconvolution of the RF signal over time comprises using the waveform of the RF signal as a frequency domain signal over time. Using the waveform of the RF signal as a frequency domain signal causes the frequency axis associated with the RF signal to change over time based on the rate at which the frequency relationship between the spectral weighting function and the optical carrier frequency is corrected over time. 13. The method of claim 13, comprising producing in. Modulating the input RF signal to an optical carrier frequency comprises simultaneously modulating the input RF signal to a plurality of optical carrier frequencies, and the spectral weighting function comprises a plurality of peaks corresponding to the plurality of optical carrier frequencies. The method according to any one of claims 10 to 14, wherein the method comprises the above. Modifying the frequency relationship between the spectral weighting function and the plurality of optical carrier frequencies comprises modifying the spacing of the plurality of optical carrier frequencies and / or the spacing of the plurality of peaks of the spectral weighting function. , The method according to claim 15. The method according to any one of claims 10 to 16, wherein modifying the frequency relationship between the spectral weighting function and the optical carrier frequency comprises adjusting an optical filter. The method according to any one of claims 10 to 17, wherein modifying the frequency relationship between the spectral weighting function and the optical carrier frequency comprises changing the carrier frequency of the laser light source. Modulating the input RF signal to an optical carrier frequency comprises simultaneously modulating the input RF signal to a plurality of optical carrier frequencies, and the frequency relationship between the spectral weighting function and the optical carrier frequency over time. Modification 10 to any one of claims 10-18 comprises transmitting the modulated optical signal to a frequency dependent delay element and applying various delays to each optical carrier frequency and the corresponding modulated input RF signal. The method described in the section. 10. It is claimed that modifying the frequency relationship between the spectrum weighting function and the optical carrier frequency over time is based on the rate of change per unit time, and calculating the RF spectrum is based on the rate of change. The method according to any one of 19 to 19. Any of claims 10-20, further comprising applying frequency modulation to the input RF signal prior to modulating the input RF signal in order to increase the spacing between the optical carrier frequency and the modulated optical signal. The method according to item 1. A method for controlling RF signal analyzers,
Generates a modulator control signal to control the modulation of the input RF signal to the carrier frequency by the optical modulator that generates the modulated optical signal, and performs a sweep over the frequency range to modify the carrier frequency and the modulated optical signal. By doing so , the frequency relationship between the spectrum weighting function characterized by the spectrum weighting module that modifies the modulated optical signal and the carrier frequency can be modified over time.
Receiving a sensed RF signal indicating the modified signal generated by the spectral weighting module over time.
A method comprising calculating the RF spectrum of the input RF signal over time based on the sensed RF signal.

Images